Adaptive Filter for Preceding Vehicle Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional inter-vehicle control systems struggle to select a preceding vehicle at a timing suited to the driver's perception across various vehicle speeds due to varying filter characteristics based on inter-vehicle distance, leading to instability and responsiveness issues.
Innovation Solution
The system includes curvature estimating, object position detecting, instantaneous probability calculating, filter calculating, preceding vehicle selecting, and inter-vehicle time calculating means, with filter characteristics adjusted based on inter-vehicle time to optimize filter calculation characteristics according to vehicle speed, ensuring suitable selection timing regardless of speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If the filter is set to be strong to suppress variations in own vehicle lane probability, then stability of preceding vehicle selection is improved, but responsiveness to merging vehicles deteriorates
Solution Approach 1:
The filter coefficient is made dynamically adjustable based on detected driving conditions. The system switches between a first filter coefficient (stronger filtering) and a second filter coefficient (weaker filtering) depending on whether the vehicle is traveling on a straight road or a curved road, allowing the filter characteristics to adapt to different scenarios rather than using a fixed coefficient
Solution Approach 2:
The system changes the filter coefficient parameter based on the curvature of the road. When the road curvature exceeds a threshold, the system switches to a different filter coefficient, thereby adjusting the filtering strength according to the road geometry to balance stability and responsiveness
2Stability of the object's composition
If the filter coefficient is changed based on inter-vehicle distance, then stability at long distances is improved, but suitability at different vehicle speeds deteriorates
Solution Approach 1:
The system changes the filter coefficient parameter based on road curvature rather than inter-vehicle distance. This parameter change strategy better reflects driver perception across different speeds, as curvature directly affects the estimated traveling road and lane probability calculations
Solution Approach 2:
The filter characteristics are dynamically adjusted according to the detected road curvature and vehicle speed conditions. The system selects different filter coefficients based on whether the vehicle is on a straight or curved path, making the filtering adaptive to the actual driving scenario
3Stability of the object's composition
If the filter becomes stronger at 80m inter-vehicle distance for low-speed cruising, then stability is improved, but responsiveness during high-speed cruising deteriorates
Solution Approach 1:
The filter coefficient is dynamically adjusted based on the combination of vehicle speed and road curvature. The system determines whether to apply stronger or weaker filtering by evaluating both speed and curvature conditions, allowing optimal filter characteristics for each driving scenario
Solution Approach 2:
Different filter coefficients are applied based on local driving conditions (curvature and speed). The system identifies specific condition ranges (e.g., low speed with certain curvature) and applies appropriately tuned filter coefficients for each local scenario rather than using a uniform filtering approach
Data Source
AI summary
A preceding vehicle selection apparatus estimates a curvature of a road on which an own vehicle is traveling, detects an object ahead of the own vehicle, and determines a relative position in relation to the own vehicle. Based on the curvature and the relative position, an own vehicle lane probability instantaneous value is determined. This instantaneous value is a probability of the object ahead being present in the same vehicle lane as the own vehicle. By a filter calculation on the instantaneous value, an own vehicle lane probability is determined. Based on the own vehicle lane probability, a preceding vehicle is selected. An inter-vehicle time required for the own vehicle to reach a detection position of the object ahead is calculated. Based on the inter-vehicle time, characteristics of the filter calculation are changed such that an effect of the own vehicle lane instantaneous value increases as the inter-vehicle time decreases.